It should do this and it's not semantically different from what happens with non-MBS disks. The log I pasted is what unmaps the volume, I am not sure why it returned successfully if the volume wasn't unmapped, if possible please attach vdsm and supervdsm logs from the relevant, perhaps there's some clue there. But we essentially use cinderlib's `disconnect`, so perhaps it hasn't errored
On Mon, Aug 23, 2021 at 10:05 AM <[email protected]> wrote: > > When I check the status of the rbd volume, watcher still exists. Wathcer is > /dev/rbd0 in the ovirt vm. > $ rbd status mypool/volume-3643db6c-38a6-4a21-abb3-ce8cc15e8c86 > Watchers: > watcher=192.168.7.18:0/1903159992 client.44942 > cookie=18446462598732840963 > > And the attachment information was also left in the volume_attachment of > ovirt_cinderlib DB. > > After manually unmap /dev/rbd0 in the ovirt vm and delete the db row, the pvc > was deleted normally. > Shouldn't those tasks be done when deleting the pod? > _______________________________________________ > Users mailing list -- [email protected] > To unsubscribe send an email to [email protected] > Privacy Statement: https://www.ovirt.org/privacy-policy.html > oVirt Code of Conduct: > https://www.ovirt.org/community/about/community-guidelines/ > List Archives: > https://lists.ovirt.org/archives/list/[email protected]/message/PSL4JPAMEQ5NHICWI34YI3HO62J2T3MB/ _______________________________________________ Users mailing list -- [email protected] To unsubscribe send an email to [email protected] Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/[email protected]/message/646F25UD5ARFWPUOYMQABSIMGD63Y7SJ/
